BaSzErr - blog:2023:11:03

2023-11-03_-_camera_light_for_print_monitoring

anonymous@undisclosed.example.com (Anonymous) — Fri, 03 Nov 2023 11:29:41 +0000

2023-11-03 - camera light for print monitoring

i use camera to record prints. the reason being, that i'm often away from printer, when it's working. yet when print fails, it's very insightful to make sure one can understand what happened in that critical moment that lead to it.

my camera setup worked nice for some years. it even has it's own IR lighting, to make sure it can see in the dark. that's all great. the problem is that due to my printer location, camera is facing window… and i'm printing on the mirror… and you guessed it – there are times of a day, where direct or reflected sun light makes camera to decrease sensitivity and then whole print is under-lighted, causing it to be too dark to actually see anything. this forced me to keep lights on when printing during day. that's sub-optimal.

having some time to spare and some spare parts from past projects, i've added 3x17cm strips of bright LEDs, to illuminate the print, from the “right direction”.

how it works? well – too good. ;) when cover was black and printing in dark filaments, it was good. however now that my hotend cover is no longer black and i often print in white PLA, there's too much reflections, making print hard to see.

so in next days i'll likely add a potentiometer with the PWM control of lighting, to make sure it can be adjusted to actual demand at hand.

2023-11-03_-_heater_pid_tuning_in_3d_printer

anonymous@undisclosed.example.com (Anonymous) — Fri, 03 Nov 2023 12:56:37 +0000

2023-11-03 - heater PID tuning in 3D printer

whenever hotend is changed, or heatbed covered / rebuild, PID needs to be tuned, to make sure it performs well (i.e.: fast heating, stable temperature, etc.).

the process is simple – just connect USB cable to your printer and run serial terminal. eg. for anycubic chiron, using picpcom that'd be:

picocom --imap lfcrlf --echo --baud 250000 /dev/ttyACM0

first check existing settings with M503. on my firmware it was (stock):

M301 P20.00 I0.50 D106.55
M304 P97.10 I1.41 D1675.16

first is heatend, while second is heatbed.

note that stock values differ a lot from values on my customized setup. this is to be expected. make sure to tune these every time you change sth around heating or cooling.

hotend PID

now run 10x tuning procedure for heatend, for 230 deg.C, and then save it to RAM (i.e. as current values – that's the U1 part at the end):

M303 E0 S230 C10 U1

my output was:

#define  DEFAULT_Kp 25.67
#define  DEFAULT_Ki 1.80
#define  DEFAULT_Kd 91.43

now write these values to EEPROM:

M500

heatbed PID

next run 10x tuning procedure for heatbed, for 80 deg.C, and then save it to RAM:

M303 E-1 S80 C10 U1

my output was:

#define  DEFAULT_bedKp 213.50
#define  DEFAULT_bedKi 42.59
#define  DEFAULT_bedKd 267.60

now write these values to EEPROM:

M500

2023-11-03_-_new_hotend_cooling

anonymous@undisclosed.example.com (Anonymous) — Fri, 03 Nov 2023 11:14:13 +0000

2023-11-03 - new hotend cooling

i use anycubic chiron 3D printer.

it got modified over time a lot, including: second surface mirror as a printbed and proper cableguides for OY axis.

recently i've decided to change hotend to full-metal one creality spider.

the main criteria for it was for it to be full-metal… and an (almost) drop-in replacement for E3D V5 used by chiron. as you might have guessed by now – the “almost” part became a theme of a last couple days or so… :/ i'll detail the story out in a separate post. for now i'll only focus on a new cooling mechanism.

the idea was to make cooling fan blowing directly on the radiator, including proper guides, to make sure air travels pass it's ribs. it took this as an opportunity to also make it as open as possible (both in OSS-spirit and space-wise ;)). last but not least – filament cooling duct was redesigned along the way.

after quite some designing and testing, i've ended up with this.

note that the breakout PCB is now on the front. this makes accessing cables so much easier than before. it also prevents cables from obstructing air flow, as in original design. one thing that could still be adjusted is the height of the PCB – it would make sense to put it 5-10mm higher, to make more space for plugs behind it.

i've printed it out in PET-G and just to be on a safe side decided to cover the bottom part with kapton tape. while it's likely an overkill (it does not touch hotend and is some centimeters away from heated bed at all times), i decided to opted out on a safe side. ;)

after mounting, there's plenty of open space on the left, to make sure warm air can escape. kapton and (fairly closed) bottom parts prevent heat from creeping in from the bottom.

last, but not least – filament cooling fan, with dict pointer slightly below the nozzle.

with that, both part cooling and radiator cooling are significantly improved over the stock version. with hotend heated up, it can be seen that bottom of the radiator is around 45 deg.C range, which is totally fine/safe even for low-temperature PLA.

as usual model is available in OpenSCAD format.

2023-11-03_-_switching_to_full-metal_hotend

anonymous@undisclosed.example.com (Anonymous) — Fri, 03 Nov 2023 12:56:51 +0000

2023-11-03 - switching to full-metal hotend

recently i was doing a lot of PET-G printing, at 250 deg.C. since it was on a short notice, i needed to print it during the day, while sitting and working in the very same room. while we'll all die one day, i'd prefer it was not PTFE-fumes-induced-cancer. after some preps and research, i've decided to switch to full-metal hotend, so that no PTFE will be anywhere near (too) hot elements. my choice was creality hotend.

for PET-G it worked great, pretty much out of the box. issues started with lower temperature filaments, like PLA and Silk. especially PLA was notoriously jamming (note: jamming – not clogging!) hotend. i'll spare you quite a bit of reading of trial-and-(mostly)-error phase and instead distill down it all to key takeaways.

problem statement

1st of all – i was not dealing with clogging here. clogging is when filament obstructs nozzle's opening (you know – the 0.4mm part ;)). usual reason is: dirt, dust or PTFE getting there. jamming is when filament gets stuck somewhere before. this is a common thing for PLA + full-metal hotends combination and this turned out the be what caused my headaches, too.

winning combo

in order to combat it, i did the following: - designed improved radiator cooling, to ensure cold-end of the hot-end stays cool at all times - used thermal-conducting paste between heat-break and radiator (note that such paste should never be used between heat-break and heating element!) - slightly tightened extruder's gear to filament - added 2 drops of vegetable oil into the hotend (from the top)

i'd like to put a special emphasis on the very last step. i realize if you're new to this concept, it does sound like a joke. i can assure you it is not! more over – i think this was the biggest contributor to solving my problems! yet since it was a 1-way route (you cannot just “undo” oiling!), it was the very last step i took.

when oiling make sure to use natural, vegetable oil, not mineral oil as mineral oil is reported to negatively affect print strength of your parts. also make sure this is an oil suitable for frying as it will be at 250+ deg.C when hotend is heated. for more details check out this video explaining in detail how to oil internal of your hotend, to make sure you get a desired effect.

after applying these changes, i'm happy to report days worth of printing time w/o a single jam. compared to previous “jam every 15 minutes”, that's a huge thing.

final notes

for the record – before doing most of the above, i was experimenting with retraction. this is often suggested way to fix jamming, as “retracting” means pulling hot, melted filament into the sticky upper part of the hotend, with not PTFE to ensure it does not stick to the walls. however my printer has a Bowden extruder, end reducing retraction to 1-2mm (that was the top i could afford before having regular jams) caused too much oozing and thus stringing, effectively decreasing print quality and increasing post-processing time. while i'm now back to previous retraction settings, i did new to shorten it significantly when doing waterproof prints, where large retractions caused objects to leak.

i consider jamming problem to be now solved. my 3D printing SDK has been adjusted with all the required changes, too. as of writing these words, SDK's changes are not yet released, yet are available on master branch already.

creality spider's hotend mounting and cooling is available as well. note that the design uses threaded inserts for screws. this makes assembly, fixes and changes a breeze!

oh – and btw. when you change anything around hotend or heatbed, make sure to tune PID settings of the corresponding element.

overall impressions

it took me a lot of time to adapt my printer and processes to the full-metal hotend change. however i'm pleased with the overall results – no more PTFE fumes! you really can tell the difference! print quality is subjectively slightly better, too.

once you know how to adapt to a new setup, i think it's a change worth contemplating.

wrt my particular choice of hotend, there's one problem i have not noticed before buying. nozzles are in odd size and are only made by creality. while these are made from hardened steel, i'd expect them to last for a very long time. however these are 10-20x more expensive, compared to classic brass nozzles i have plenty of in my closet.